U.S.

Department of Commerce Bureau of Standards

RESEARCH PAPER RP688

Part of Bureau of Standards Journal of Research, vol. 12, June 1934

PREECE TEST (COPPER-SULPHATE DIP) FOR ZINC

COATINGS
E. C. Groesbeck and H. H. Walkup

abstract
The Preece test is in common
use for inspection purposes to determine the
uniformity of galvanized coatings, even though the results are sometimes erratic.
An experimental study made with wires with different types of zinc coating and
with wires consisting of commercially pure zinc and iron indicates that the pH
value of the copper sulphate solution, the rate of solution of the zinc coating, and
the cleanness of the test specimens are important factors in the Preece test. The
use of cupric hydroxide instead of cupric oxide for neutralizing the copper sulphate
solution is recommended, since a stable pH is much more quickly reached by its
use. Electroplated and sprayed zinc coatings dissolve more rapidly in the copper
sulphate solution than do hot-dipped or galvannealed coatings. Pure zinc
coatings showed, during their solution, a "potential minimum" (with respect to
a copper wire) which was practically absent with coatings that contain iron,
such as on galvannealed wires and the inner layer of hot-dipped galvanized wires.
This minimum may be caused by the delayed coagulation of basic zinc compounds
which coagulation may be accelerated by the presence of iron. The causes and
remedies of premature and delayed endpoints are explained. A method for
measuring the thickness-distribution of the zinc coating by the use of a suitably
acidified copper sulphate solution is suggested.

CONTENTS
Page
I. Introduction 786
II. Preece test _ _ 786
1 Regular procedure 786
2. Difficulties encountered 787
3. Principle of the method 787
III. Experiments 788
1. Materials used 788
2. Factors that may affect the results 788
(a) Rate of solution of the coating 788
(1) Composition of copper sulphate solution 789
(2) Temperature 790
(3) Stirring 791
Coating tested
(4) 791
Preparation 791
Composition and structure 792
(b) Premature endpoints 793
(1) Plastic films 793
(2) Potential measurements 794
(3) Oxygen concentration cells 798
(c) Delayed endpoints 799
(1) Adherence of copper 799
Effects of time and pH 799
Effects of composition and tempera-
ture 800
Use of bare steel as endpoint 800
(d) Distribution of thickness 800
IV. Conclusions- : 801
V. Selected references 802
785
786 Bureau of Standards Journal of Research [Vol. it

I. INTRODUCTION
Zinc coatings find wide use for protecting iron and steel against
corrosion, because zinc is more readily corroded than iron when they
form a "couple" or electrolytic cell in the presence of a conducting
liquid. In such a cell the zinc is the anode or dissolving electrode.
Since this method of protection depends upon the protective value of
the gradual corrosion of the zinc the protective value of the zinc
coating depends primarily on its thickness at the thinnest spot.
The total weight or average thickness of a zinc coating is determined
by some sort of a stripping method. A
common stripping method
consists in the removal of the coating by hydrochloric acid containing
antimony chloride. If a determination of the distribution of the
coating and its minimum thickness is desired, use may be made of
a solution that will dissolve the zinc coating at a uniform rate and will
reveal the base metal (iron) as soon as it is exposed. copper A
sulphate solution of definite concentration is widely used commercially
for this purpose in the procedure known as the "Preece test" (l), 1
Although this test is useful, experience has shown that the results
are not always reproducible and consistent, (2), (3), (4), (5), espe-
cially when different types of zinc coatings are tested.
The experiments reported in this paper were made to determine
relationships rather than to secure absolute data. This policy was
prompted by a desire to establish the cause and possible remedies for
the observed difficulties rather than the degree of error produced by
any one variable.
II. PREECE TEST
1. REGULAR PROCEDURE
The Preece test is applied by dipping the zinc-coated article, which
has previously been cleaned from grease or other foreign material,
into a "neutral" solution of copper sulphate of a specified concentra-
tion (1.27 mole). This concentration is usually denned by its specific
gravity (1.186 at 18 C), and the solution is maintained at a tempera-
ture of 18 C during the test. (6) The sample is immersed in the solution
for 1 minute, removed, washed in running water, and freed from the
loosely precipitated copper by light rubbing. The 1 -minute immer-
sions and cleanings are repeated until an endpoint is obtained in the
form of a bright adherent deposit of copper. This indicates that the
iron has been exposed there. The number of immersions (dips)
reported for each sample is one less than that required for the appear-
ance of adherent copper.
The use of this test in specifications rests on three assumptions:
(a) The zinc coating dissolves at a uniform rate; (b) adherent copper
is deposited as soon as the underlying iron is exposed; and (c) no
adherent copper appears until the zinc coating is entirely removed at
that point. Under such conditions the ability to withstand a speci-
fied number of immersions is evidence that the zinc coating possesses
a certain minimum thickness.
1
Figures in parentheses here and throughout the text refer to the numbers used in the list of references
at the end of this publication.
waiSif] Preece Test for Zinc Coatings 787
2. DIFFICULTIES ENCOUNTERED
Practically, these criteria are approximately met by pure zinc
coatings, such as those produced by plating, and fairly well by hot-
dipped galvanized coatings. However, with coatings produced by
" galvannealing " 2 or by sherardizing 3 the results are less consistent
and are not always proportional to the thickness of the coatings.
Such discrepancies may be the result of (a) differences in the rate of
solution of the zinc coatings, (b) the formation of adherent copper
before the zinc coating is all dissolved at that point that is, a pre- —
mature endpoint, or (c) the absence of adherent copper on iron that is
exposed — that is, a delayed endpoint.

3. PRINCIPLE OF THE METHOD

At first glance, it may appear that the method depends merely on
the displacement of copper by zinc, according to the reaction.

[1] Zn + CuS0 4 = Cu + ZnS0 4

This reaction, however, does not account for the deposition of adher-
ent copper on the exposed iron rather than on the undissolved zinc.
It is much more probable that the reaction is electrolytic that is, —
solution of zinc occurs at one point comprising an anode, with the
deposition of copper at another point serving as a cathode. The cell
may then be written:
[2] Zn (anode) |
CuS0 4 |
Cu (cathode)
or [3] Zn |
Cu++ |
Cu
This formula represents the system before any reaction occurs.
However, after any finite period some zinc is in the solution adjacent
to the zinc surface, and the cell is then represented by

[4] Zn |
ZnS0 4 |
CuS0 4 |
Cu
or [5] Zn |
Zn++ |
Cu++ |
Cu
In any case, the chemical reaction is

[6] Zn + Cu ++ = Zn+ + + Cu
To produce an electrolytic action, there must be a difference in
potential between the two points which become, respectively, anode
and cathode. This difference may arise from various causes and
differences in the composition or temperature of the adjacent solu-
tion. According to this theory, a homogeneous piece of pure zinc
will not react with a homogeneous solution of copper sulphate (just
as very pure zinc dissolves very slowly in pure sulphuric acid).
The fact that a portion of a zinc specimen or of the zinc coating
on steel does replace copper from a copper sulphate solution at once
upon immersion is evidence that small areas of the zinc have slightly
more positive single potentials than other areas, which latter serve
as cathodes upon which the copper is deposited by the dissolving of
the adjacent anodic areas. As soon as any copper is deposited,
however, it serves as a cathode to the adjacent zinc, and the latter
2 A proprietary process, in which the article (usually wire) is first coated with zinc by hot-dipping and is
then heated.
3 A zinc cementation process, in which the article is coated by heating in contact with powdered zinc.
788 Bureau of Standards Journal of Research [Vol. is

dissolves with an accompanying precipitation of copper on the already

deposited copper. This copper is nonadherent, since the surrounding
and underlying zinc is continually dissolving. When a zinc coating
on iron is completely dissolved at any point and the underlying iron
is thereby exposed, the latter serves as a cathode to the surrounding
zinc. An adherent deposit of copper is produced and constitutes
the true endpoint. It should be noted that the production of ad-
herent copper on the iron requires the presence of zinc adjacent to it.
If a clean piece of iron is dipped into a neutral copper sulphate solu-
tion, the deposited copper is not very adherent, though usually more
so than is copper deposited directly on zinc.

III. EXPERIMENTS
1. MATERIALS USED
Although committee A-5 of the American Society for Testing
Materials expressed the opinion that the Preece test is unreliable and
should be abandoned as a basis of specifications for galvanized sheet
and plate, it admitted that this test is a quick and simple method for
inspecting galvanized wire, for which it gives comparable results if
properly applied (4). Since galvanized wire and wire products are
important, wires with various types of zinc coatings were used in
this study.
Wires coated with zinc by the hot-dip galvanizing, galvannealing,
and electroplating (electrogalvanizing) processes were obtained from
certain manufacturers. Two lots of hot-dipped galvanized wire
from different sources (designated as c and k) were used. The
electroplated wires were plated in a highly acid bath at a high-
current density (about 750 amperes per square foot) with insoluble
anodes. Wires were coated with zinc by metal spraying in the
laboratory. As sherardizing is not commercially applied to wire,
this type of coating was not included. Soft steel and commercially
pure iron and zinc in wire form were used in the study of the effects
of certain factors. The history of the iron, steel, and zinc wires is
not known. The test specimens were generally about 3 inches (7.5
cm) long and % inch (3 mm) in diameter.
The term "Preece solution" used throughout the text refers to a
copper sulphate solution of the concentration specified by the Ameri-
can Society for Testing Materials (6) regardless of its other charac-
teristics, such as pH.

2. FACTORS THAT MAY AFFECT THE RESULTS

With the above theory as a basis, an effort was made to predict
and to test the major effects of variations in operating conditions
upon the time of formation, physical properties, and adherence on
steel of the precipitated copper.

(a) RATE OF SOLUTION OF THE ZINC COATING

The use of the Preece test is generally based upon the assumption
that the rate of solution of the zinc in various galvanized materials
is practically the same. But this rate of solution depends upon
1, the composition of the copper sulphate solution; 2, its tempera-
Groesbeck]
Walkup J
Preece Test for Zinc Coatings 789

ture; 3, the stirring; and 4, the composition and structure of the zinc
coating, including any surface films that may be present initially or
that formed during the reaction.

(1) COMPOSITION OF COPPER SULPHATE SOLUTION

The employed has a specific gravity of 1.186

solution generally
at 18 Cwhich corresponds to a concentration of 1.27 mole or
(6),
315 g/1 of CuS0 4 -5H 2 0. Although this concentration is artibrary,
it is not very critical, and hence no exhaustive tests were made on
the effect of concentration on the rate of solution of the zinc. In
general, an increase in copper content accelerates the solution of the
zinc, and vice versa. The practical significance of this effect is that
only a limited amount of a zinc coating should be tested in a given
volume of copper sulphate solution, because the precipitation of
copper reduces the concentration of copper sulphate.

4.0

3.5
r
l
— °*-— .,
A

-I B
'|

3.0
A- CuO POWDER
B-CU(OI0 2
•

PS 1 1 I 1
1

20 40 60
TIME IN CONTACT WITH POWDER , HOURS
Figure 1. Change in pH value of the Preece solution with different neutralizing
agents on standing.

The neutralizing agents were added to portions of an unneutralized Preece test solution and the pH value
was determined at intervals.

The neutral solution specified in the test is generally obtained by

treating the solution of the salt with an excess of cupric oxide at room
temperature. Small variations in acidity are most conveniently ex-
pressed in terms of hydrogen-ion concentration or pH. neutral A
solution of copper sulphate is "acid" because of hydrolysis— that is,
it has a pH less than 7, the neutral point of water. No published
data could be found on the exact pH of copper sulphate solutions of
about the concentration used in this work. However, the results of
numerous measurements made with a quinhydrone electrode show
that the equilibrium pH is about 3.3, which defines a truly neutral
solution of copper sulphate.
The addition of finely divided cupric oxide to a copper sulphate
solution with a pH of about 2.7 was found to raise the pH to nearly
4 in a short time (fig. 1). The pH then decreased to 3.3 on long
standing (6 weeks) and an insoluble substance, believed to contain
basic cupric sulphates (7), (8), was precipitated. It is evident that
the final pH is the result of an equilibrium of sulphuric acid with
790 Bureau of Standards Journal of Research [Vol. 12

cupric hydroxide or basic cupric sulphate and not with cupric oxide.
It is surprising that cupric oxide, which is not easily soluble, would
have " over-neutralized" the solution. The use of basic cupric car-
bonate in place of cupric oxide gave similar results. On the other
hand, cupric hydroxide very rapidly produced a pH value of about
3.3, which remained constant for a long period (fig. 1). The use of
either powdered copper hydroxide or basic copper carbonate has been
previously recommended for neutralizing purposes. (9)
Since, as will be shown, the pH of the solution affects the results,
it is important to reach the desired pH quickly and to maintain it
during the test. The substitution of cupric hydroxide for cupric oxide

600--

500

<
o
H 200 +

Figure 2. Effect of
12
pH
pH OF CuS04 SOLUTION
3

on the rate of solution of a galvannealed coating.

4

In each experiment the coated wire was subjected to the Preece solution at about 20 C for successive
periods of 30, 60, or 90 seconds, respectively, until a true endpoint was obtained. The pH of the solu-
tion was varied by suitable additions of H2SO4. The specimens were cleaned with tap water and a dry
towel after each immersion.

in neutralizing the solution appears very desirable. As noted above,

pH values greater than 3.3 may be obtained with cupric oxide or
carbonate, but such solutions are unstable. For experimental pur-
poses, sulphuric acid may be added to produce values below 3.3, pH
which can conveniently be measured with a quinhy drone electrode.
The rate of solution of the zinc was generally accelerated (fig. 2), by
increasing the acidity (decreasing the pH) of the solution.

(2) TEMPERATURE
The rate of solution was found to increase, as expected, with the
temperature (fig. 3). The reproducibility of these data does not
Oroesbeck]
Walkup j
Preece Test for Zinc Coatings 791

warrant the computation of exact temperature coefficients, but they

show that the rate of solution is doubled by an increase of about 15 C.
This indicates that to obtain an accuracy of about 10 percent in the
Preece test it is necessary to keep the temperature constant (for
example, at 18 C) within 2°.

(3) STIRRING

Moving the specimen, or agitating the solution, has the same effect
as increasing the concentration of copper sulphate or of sulphuric
acid, because stirring replenishes these compounds close to the metal
surface. As it is impracticable to define a rate of agitation, especially
for different shapes, it is desirable to avoid any agitation.

0.35
O-HOT-DIPPED .CURVES B & E
V-ZINC SPRAYED , CURVE A
X -ELECTROPLATED C
••
,

A- GALVANNEALED , "

h- 0.2 5 -

X
o
U
z
«n 0.15
o

0.05
10 22 28
TEMPERATURE °C.

Figure 3. Effect of temperature of solution on average loss in weight for each

immersion for different kinds of zinc coatings.

(4) COATING TESTED

Preparation. —
Obviously, any metal coating should be chemically
clean when tested for its thickness or behavior. It is also obvious
that the method of cleaning should be such that it does not remove
any significant amount of the coating. Grease is best removed by an
organic solvent, such as benzene, since alkaline solutions may attack
the zinc. As it is not practicable to remove any oxide or carbonate
films with acid, mechanical cleaning must be used. The data in
table 1 indicate the greater rate of solution of zinc from a surface
cleaned with fine emery paper (for example, aluminum oxide paper no.
400). The average amount of zinc removed by the emery paper was
0.04 oz/ft 2 with a maximum of 0.07 oz/ft 2 These losses represent not
.

55948-34 9
792 Bureau of Standards Journal of Research [Vol. n

more than 7 percent error in samples with coatings of 1.0 oz/ft 2 of the
surface, but the error for thinner coatings would be correspondingly
larger. It is thus evident that cleaning with even a fine abrasive,
such as emery, must be done with great care.

Table 1. Effect of the initial cleaning method on the loss in weight during the
first immersion

[The solution was prepared by dissolving 36 g CuS04-5H 2 to 100 ml H2O, and adding a slight excess of
Cu(0H>2. It was then shaken, let stand for 30 minutes and filtered. The temperature of the solution
was approximately 20 C, and the immersion periods were 60 seconds]

Loss in weight during first

immersion

Cleaning method
Electro- Hot- Galvan-
plated dipped nealed
wire wire (k) » wire

Oz/fV Otlfl* Oz/ft*

0.155 0.137 0.128
.199 .165 .152

» See sec. Ill, 1.

—
Composition and structure. The various types of zinc coating used
in this study vary in both composition and structure. Electro-
plated coatings consist of practically pure zinc (except when mercury
is used in zinc cyanide baths, in which case the deposit may contain
up to 1 percent of mercury). The outer surface of hot-dipped coatings
consists of nearly pure zinc, while the portion next to the steel is
composed of a zinc-iron alloy. A galvannealed coating consists
throughout of zinc-iron alloys.
As zinc has a greater solution pressure than iron, it is reasonable to
assume that zinc will dissolve more rapidly than a zinc-iron alloy.
This is confirmed by the data in table 2, which shows that the rate of
solution of the coating and the number of immersions corresponding
to a given weight or thickness of zinc are not uniform. Hence, the
results for different types of coating are not directly comparable.
They may be correlated in either of two ways: 1, by specifying a differ-
ent number of immersions for a given thickness of each type of coat-
ing, or 2, by changing the immersion period so that the number will
be the same for a given thickness. For example, if the period for
electroplated coatings were retained at 60 seconds, the other periods
would have to be approximately 55 seconds for zinc-sprayed, 80 sec-
onds for hot-dipped, and 120 seconds for galvannealed. These
figures are, however, not applicable in a practical way, because the
weight of coating dissolved is not exactly proportional to the time.
It is much simpler to use 60-second immersions and to specify the
number of immersions that correspond approximately to any desired
thickness of a given type of coating.
The initial increase in the solution rate reported for the hot-dipped
specimens, followed by a decrease (table 2), is in agreement with
Bablik's observation (5) that the solution rate for hot-dipped
galvanized sheets increases at first and then decreases with continued
immersions.
Groesbeckl
Walkup J
Preece Test for Zinc Coatings 793

Table 2. Loss in weight of commercially pure zinc wire and of different types of
zinc coating on iron wires for successive immersions in the Preece solution

Average weight
Material
Immersion Loss in of coating, by
number weight stripping
(HCl-SbCls)

Oz/p Oz!P
fl _ . 0.170
2 .184
{3 .186
4 .193
15 .163

Total ._ .90
.18

fl .155
2 .247
3 .245
1.32
4 .226
5 .224
6 . .215

Total 1.31
Average.. _ .. . ... . .22

fl .137
2 .208
Hot-dipped wire (k) ° . ______ \ 0.79
3 .168
u___ .129

Total .64
Average.. .... . . .16

fl— .128
2___ .106
3 .104
4 .110
Galvannealed wire ... . 1.04
5 _ .120
6 .112
7 . .104
8___ .108

Total .89
.11

fl. . .243
J2 .251
Zinc-sprayed .. 6 1. 17
13 .243
[4 .220

Total .96
Average ._. . .24

« See sec. Ill, 1.

* The weight of coating was determined by the difference in weight of the specimens before and after
the metal-spraying process. Solution temperature and immersion periods as in table 1.

b. PREMATURE ENDPOINTS
(1) PLASTIC FILMS

Data on the rate of solution of zinc coatings are of practical value

only if the endpoint can be reliably detected. The appearance of
adherent copper on pure zinc coatings is usually definite and repro-
ducible, but adherent copper sometimes appears on hot-dipped or
galvannealed coatings before the coating has entirely dissolved under-
neath the copper deposit.
During cleaning, particles of precipitated copper may adhere to a
zinc surface by being squeezed into recesses in a rough surface or into
a plastic film, which may be formed during the reaction. The former
result is unusual, but may occur if the pH value is very low (acidic),
in which case the zinc is more violently attacked and consequently
may be roughened. The occurrence of this latter result is favored by
794 Bureau oj Standards Journal of Research [Vol. is

conditions which permit the formation of basic precipitates on the

zinc surface. The most obvious cause is too high a value of thepH
solution. Under such conditions there will be a tendency for the
precipitation of basic compounds from easily hydrolyzable salts. At
the cathode of an electrolytic cell, the pH value tends to increase
whenever hydrogen is evolved, a reaction that may occur when zinc
is immersed in a copper sulphate solution which necessarily contains
hydrogen ions. The hydrogen-ion concentration will also be reduced
if any dissolved oxygen is electrolytically reduced on a cathodic area.
Exact data are not available to permit the predication of the pH value
at which basic compounds of copper, zinc, and iron may form in
these solutions, especially as the respective concentrations of these
metals are always changing.
If the copper sulphate is originally 1.27 mole, a complete reaction of
this solution with the zinc coating should produce a zinc sulphate
solution of the same molar concentration. Actually, some copper
sulphate is always present in the solution adjacent to the zinc surface
as a result of diffusion and convection. Furthermore, ferrous sulphate
may be present in the portion of the solution next to the metal surface
if the coating contains iron. The oxidation of the ferrous sulphate
to the ferric condition would probably be prevented by the presence
of zinc, although it may occur at a short distance from the metal
surface. Therefore, the solution next to the metal contains, shortly
after the immersion of the zinc-coated specimen, hydrogen, cupric,
zinc, ferrous, and possibly ferric, ions. If, now, some of the hydrogen
ions are removed, hydroxides (or basic salts) of copper, zinc or iron
may be precipitated along with the metallic copper, and may cause
the latter to adhere to the surface, even though no steel is exposed.

(2) POTENTIAL MEASUREMENTS

An effort was made to follow the course of the reaction when a zinc-
coated specimen is immersed in a copper sulphate solution by making
simple potential measurements in a cell consisting of (a) zinc or zinc-
coated iron wire, (b) copper sulphate solution, and (c) copper wire.
The total cell voltage was measured at intervals by means of a high
resistance voltmeter.
When a zinc wire or a pure zinc coating (curve A, fig. 4) is immersed
in the copper sulphate solution, there is a rapid decrease in potential
to a sharp minimum, followed by a relatively slight rise, after which
the potential decreases gradually. A
potential-minimum was not
observed in the case of galvannealed coatings.
There is no completely satisfactory explanation of a curve such as
A. The potential measured is that of a primary cell such as the
following:
Anode Cathode
Zinc (or zinc coating). ZnS0 4 CuS0 4 Copper
to 1.27 mol 1.27 to mol
As this cell is discharged, the concentration of zinc sulphate increases
and that of copper sulphate decreases. Both of these changes cause
a rapid decrease in cell potential, which is represented by the first,
i.e., descending, portion of the curve.
Groesbeck^
Walkup J
Preece Test for Zinc Coatings 795

If now the pH
of the solution adjacent to cathodic areas increases,
as explained above, there is a tendency for the formation of hydroxides
or basic salts. At first these may be in colloidal solution and require
a finite time for coagulation. If the latter occurs it would slow up the
reaction, retard the precipitation of copper, and permit the original
concentrations to be restored by diffusion or convection.
The absence of a potential minimum, when iron is present, may
be an evidence that the iron compounds favor coagulation and thus
prevent the " supersaturation " with basic compounds.
This explanation is consistent with curve B, which shows that the
rate of reaction, i.e., the change in potential is at first rapid and then
changes at a nearly uniform slower rate.

1.25

1.00

0.75

o
a. 0.50

ZINC WIRE
0.25 B-GALVANNEALED WIRE
b C- "POTENTIAL MINIMUM

3
TIME OF IMMERSION , MINUTES
Figure 4. Change in potential with time of immersion for zinc wire and for
galvannealed wire.
The curves are representative" of data obtained by using different galvanized and zinc wire specimens
as the anode and various CuS04 solutions as the electrolyte in the cell, specimen solution copper wire,
| |

with the electrodes placed about 2 cm apart. A 25,000 ohm voltmeter, provided with a to 2.5 volt scale,
and a stopwatch were used.

A somewhat similar curve was obtained by Kreuchen (10) for the

potential of a single zinc crystal in normal zinc sulphate solution. He
reported that polycrystalline zinc showed a slightly higher potential
without any such minimum and attributed the difference to the varia-
tion in solution pressure of zinc from a single complete crystal plane
and from a broken surface. Such an explanation does not apply to
the zinc coatings, which certainly are polycrystalline.
The effects of various factors upon the position of the potential
minimum were next investigated. The time required to reach the
minimum potential increased as the acidity of the copper sulphate
796 Bureau oj Standards Journal of Research [Vol. 1%

solution, which was at about 20 C, was increased (pH was decreased

fig. 5). as explained above, the occurrence of the minimum po-
If,
tential is related to the coagulation of hydroxides or basic salts in the
copper sulphate solution, a decrease in the of the solution would pH
increase the time needed for this coagulation to take place.
Figure 6 shows the effect of the temperature of the copper sulphate
solution on the time required to reach the potential minimum. A
satisfactory reason for the marked increase in the time, at 40 to 60 C,
for zinc, and at lower temperature ranges for hot-dipped and galvan-
nealed coatings, cannot be given, although the shifting of the peak in
the curves to the left is apparently related to the iron content of the
coating. That the presence of iron in the coating retards or prevents
the occurrence of a potential minimum was confirmed by the results
obtained (fig. 7) for a hot-dipped galvanized coating, the successive
layers of which (starting from the outside) contain increasing propor-
tions of iron. The results for galvannealed wire (curve B, fig. 4) are
also in agreement with this conclusion.

6
(0

'80- >x>

6
u. o
-°"60-
X o

Id
V o
O40- o o

a \o
Id
Q
Id
^^**^*~2
£20-
o
Id

1— 1
t_
Q I 2 3
PH OF CuS04 SOLUTION

Figure 5. Effect of pH of solution on the time required to reach the minimum

potential (b, fig. 4), when using zinc wire specimens.

If then the basic films form more rapidly on the pure zinc than on
the zinc-iron alloy, it is reasonable to suppose that the films formed on
the latter would be more adherent and hence be more likely to retain
copper particles after the cleaning operation.
That plastic material may be formed on the surface of the coatings,
especially of zinc-iron alloys, was shown by numerous observations,
summarized (a, b, c), as follows, in which the specimens were washed
in running tap water and wiped dry with a towel after immersion in
copper sulphate solution.
(a) A pure zinc wire acquired a dull color after a short immersion
and a gray color after a period corresponding approximately to the
minimum, b, in figure 4. A longer immersion produced a bright
metallic luster, and little or no plastic material could be detected on
the surface.
Qroesbeckl
Walkup J
Preece Test for Zinc Coatings 797

(b) A plastic film was left on hot-dipped galvanized wires, the

amount of which increased with the greater iron content of the lower
layers of the coating.
(c) Plastic material was found on the surface of the galvannealed
wires.
No potentital minimum was observed when a zinc wire was im-
mersed in a copper sulphate solution containing potassium dichromate,
but the surface was coated with a black film. Immersion for a long

300

D- GALVANNEALED , CURVE A
O- HOT-DIP , " B
X-ZINC , » C

40 60 100
TEMPERATURE, °C
Figure 6. Effect of temperature of the CuS0 4 solution on the time required to
reach the "potential minimum" (6, fig. 4)-
A Preece solution of pH 3 was used.

period produced a plastic residue, which contained a white gelatinous

material (presumably basic zinc sulphate), metallic copper, and a black
substance (probably containing copper oxide). The dichromate is an
oxidizing agent that is readily reduced by zinc, and hence the pH of
the solution was raised to a point where basic compounds were pre-
cipitated, as predicted above. It was found that by immersing
galvannealed wires in a copper sulphate solution with a pH between 1
798 Bureau of Standards Journal of Research [Vol. m

and 2 for periods of 120 seconds or more the tendency for the produc-
tion of premature endpoints was decreased.

(3) OXYGEN CONCENTRATION CELLS

It was predicted that copper may deposit adherently on zinc sur-
faces if there is sufficient difference of potential to make certain areas
cathodic. Such a condition may arise if a portion of the zinc surface
is either (a) protected by a film that reduces its solution pressure, or
(b) exposed to a copper sulphate solution having a higher oxygen
content than that in contact with another portion of the surface. It
was found that the first cause can be eliminated by carefully cleaning

1.25

i.oo

0.75
*-*^
LU
H
O
a. 0.50
_i

HI
o LETTERS ON CURVES REFER TO IRON
CONTENT OF EXPOSED SURFACE
0.25 A- LOW
B-INTERMEDIATE
C-H.IGH

3
TIME OF IMMERSION, MINUTES

Figure 7. Effect produced by the iron content of the exposed surface of hot-dipped
zinc coatings on the potential-time relation.
Specimens of hot-dipped galvanized wire were immersed in a C11SO4 solution for different periods of
time and were cleaned by rubbing with a pencil eraser until a bright surface was obtained They were then
.

immersed in a Preece solution of pH 3.2 which was at 10 C for securing the data given.

the specimen before the first immersion and by rubbing it rather

firmly between immersions, for example, with a soft pencil eraser.
The effect of an oxygen concentration cell is noticeable near the
"water line" of the Preece solution, and it may be accentuated by
rotating the partially immersed specimen eccentrically, whereby
localized aeration is produced. If this effect is combined with local
surface contamination, copper readily deposits on the uncleaned part
of the surface. Agitation of the specimen in the Preece test should,
therefore, be avoided.
If, for any reason, the zinc dissolves more rapidly at one point than
at another, for example, at the water line or near previously exposed
steel or abraded areas, copper will deposit on the steel there sooner
Groesbeckl
Walkup J
Preece Testjor Zinc Coatings 799

than at other points. Strictly speaking, this latter effect is not a

premature endpoint, but is a true criterion of a defective coating.
Such defects, if incidental to the preparation or testing of the speci-
mens, should, of course, be disregarded.
(c) DELAYED ENDPOINTS
(1) ADHERENCE OF COPPER
Effects of time and pH. —
If, for any reason, copper does not adhere
to steel that is exposed in the test the result will be misleading. Be-
cause of the difficulty of measuring or expressing the degree of ad-
herence, the following empirical procedure was employed. copper A
coating was considered adherent to the steel when it could not be
removed by vigorous rubbing with dry cotton or by flaking with the
finger nail. Tests were made to determine the length of time speci-
mens could be exposed to the copper sulphate solution before the
120

O
LJ
to 80

s-a

2
O
CO

" 40 A-GALVANNEALED , GUF?VE A

D -HOT-DIP ; » B
O-SOFT STEEL , •• C
X-IRON . " C

0.5 1.5 2.0 2.5 3.0 3.5 4.0

pH OF CuS0 4 SOLUTION
Figure 8. Effect of thepH value of the C11SO4 solution on the " adherence time'
of the copper deposited on iron and steel.

copper deposit became detached from the iron or steel, which had
been initially exposed by filing off a portion of the coating.
In the experiments reported in figure 8, several specimens of the
same material were immersed for different periods, each specimen
being used only once. The curves present the observed relations
rather than the specific data. The specimens were cleaned by first
rubbing with aluminum oxide paper no. 400 and then by immersing
in a 10 percent H
2 S0 4 solution for 5 to 10 seconds, followed by washing

in water and wiping dry with a cotton swab. Immediately after

drying, they were immersed in the CuS0 4 solution at 20 C. The
degree of adherence subsequent to the immersion in CuS0 4 solution
was approximately defined by the above rubbing and flaking tests.
In all cases the adherence of the copper decreased as the pH was
increased (fig. 8), though it was adherent for approximately 1
800 Bureau of Standards Journal of Research [Vol. 12

minute when the pH was 3.3, the recommended value. On the other
hand, the copper deposited on uncoated iron or steel wire was prac-
tically nonadherent if the pH was much above 1. (This agrees with
the general use of an acidified copper sulphate solution to produce
copper films on steel by immersion.)
It is of interest to note here that a ''contrast in color" method is
made use of in a Navy Department specification (11), in which a
bright adherent copper deposit on steel is prepared each day of test
as a standard color sample for comparison with the color of the copper
deposits formed on galvanized wires in the Preece test.
Effects of composition and temperature. —
Other tests showed that
the adherence of the copper on the galvanized wires was decreased by
very small concentrations of chloride or nitrate in the copper sulphate
solution. The adherence was increased by cooling the solution, for
example, from 30 to 5 C. A decrease in concentration of copper
sulphate also increased the adherence of the copper. The latter two
effects are the results of decreasing the rate of reaction and hence,
the rate at which the copper is deposited.
Miscellaneous observations showed that the copper was not as
adherent on a perfectly clean steel surface as on a slightly tarnished
one. Evidently the clean steel was itself attacked by the solution.
When the immersions were continued beyond the usual endpoint,
the copper that was at first adherent on zinc-plated specimens became
nonadherent, while that on the galvannealed specimens remained
adherent. Adherent copper did not always form in the first immer-
sion on the steel bared by filing an electroplated wire, but did in sub-
sequent immersions. If the zinc coating is very uniform so that the
entire coating is dissolved in a given immersion (especially the first
one), copper does not readily deposit on the steel, even in the succeed-
ing immersion.
—
Use of bare steel as endpoint. In view of the uncertainty as to
whether or not copper will at once deposit on exposed steel, the ap-
pearance of bared steel may sometimes serve as an endpoint, which
is usually observed one immersion before the copper would normally
deposit. Since the appearance of steel is not so evident as that of
adherent copper, it is advisable to give the specimen another immer-
sion to see whether or not the previously exposed steel is coated with
adherent copper. In effect, this practice would not differ from the
usual procedure, except that the appearance of steel would serve as a
warning that the endpoint would probably occur on the next
immersion.
(d) DISTRIBUTION OF THICKNESS

While the Preece test is commonly used merely to detect areas

having a minimum thickness, it may be modified so as to give an
approximate idea of the thickness distribution. This is accomplished
by noting the relative areas coated with copper in immersions succeed-
ing the endpoint. The neutral copper sulphate solution (pH-3.3) is
not satisfactory for this purpose, since exposure to succeeding im-
mersions causes the previously adherent copper to become detached.
If, however, sufficient sulphuric acid is added to lower the pH to
about 0.6, for example, 3 ml of concentrated H
2 S0 4 to 100 ml of
neutral CuS0 4 the copper deposit remains adherent throughout the
,

successive immersions. The results (table 3) show that in these tests

Groesbeck']
Walkwp J
Preece Testjor Zinc Coatings 801

the plated and galvannealed coatings were more uniform than the
hot-dipped or sprayed ones.

Table 3. Proportion of area covered with adherent copper after successive immer-
sions for 30-second periods

Surface covered with deposited copper in immersion periods

subsequent to the first appearance of bared steel
Type of coating

1 2 3 4 5 6

Percent Percent Percent Percent Percent Percent

Electroplated . .. . .-. 60 95 100
5 90 99
Hot-dipped (k)° ... _. . .. 2 90 95 99
Hot-dipped (c)° -. _. 10 30 70 90 100
Zinc-sprayed- ..-.- 5 40 60 90 99 99

See sec. 3, 1.
Solution: 3 ml of concentrated H2SO4 were added to 100 ml of a solution that was prepared in the same
way as for table 1; temperature of the solution, 18 C; concentration, 1.27 mol-CuS04.5H 2 0; and pH, 0.65.

IV. CONCLUSIONS
1. If a copper sulphate solution (1.27 mole) is neutralized with
cupric oxide, a pH of 4.0 may be rapidly obtained. On long standing
the pH decreases to about 3.3, and a basic precipitate forms. A
similar result is secured by neutralizing with basic cupric carbonate.
Treatment with cupric hydroxide produces inside of 24 hours a solu-
tion with a pH of 3.3, which is permanent and probably represents
the true equilibrium pH. The use of cupric hydroxide as a neutral-
izing agent is recommended.
2. The specimens to be tested should be thoroughly cleaned before
and between immersions (dips), as otherwise adherent copper may
form on portions of the zinc surface that are not entirely clean.
Grease that is initially present may be removed with an organic
solvent, while corrosion products should be carefully removed with a
very mild abrasive.
The rate of solution of zinc coatings in copper sulphate solution
3.
is increased by raising the concentration of (a) copper sulphate or
(b) free acid. Stirring also increases the rate of solution of the coat-
ing, and therefore should be avoided in the test. Adherent copper
may form on zinc coatings if the wire specimens are rotated eccen-
trically, which probably produces oxygen concentration cells.
4. Electroplated and sprayed zinc coatings dissolve more rapidly
than hot-dipped or galvannealed coatings. For a given thickness of
coating, a greater number of 60-second immersions, or else longer
immersions must, therefore, be specified for the latter two coatings.
5. The potential of zinc, or of a pure zinc coating, in a copper sul-
phate solution decreases rapidly for about one minute to a potential-
minimum, then increases slightly, and finally graduaUy decreases. No
such potential minimum is observed with galvannealed coatings or
with the iron-rich layer of hot-dipped coatings. The minimum may
be caused by the delayed coagulation of basic zinc compounds. The
presence of iron may accelerate this coagulation to form a more dense
film, that prevents such a minimum. The position of the potential
minimum is affected by the pH and temperature of the copper
sulphate solution.
802 Bureau of Standards Journal of Research [vol. 12

6. Premature endpoints, that is, adherence of copper before the

underlying steel is exposed, may be caused by the formation of plastic
material on the metal surface during the test. This is more likely to
occur with zinc coatings containing iron, such as galvannealed coat-
ings or the inner layer of hot-dipped coatings. Its effect can be
eliminated by a thorough rubbing of the dried specimen between
immersions with a soft abrasive, for example, a pencil eraser.
7. Delayed endpoints may result from the failure of copper to de-
posit immediately on steel that is exposed. In such cases the appear-
ance of the bare steel may serve as an approximate endpoint, which
may usually be confirmed by the deposition of copper in the succeed-
ing immersion.
8. The adherence of copper to the bare steel may be improved by
lowering the pH, though it is usually adherent for at least 1 minute
when the pH is 3.3.
9. By using a more acid solution, for example, pH = 0.6, successive
immersions after the first appearance of adherent copper may be used
to study the thickness distribution of the coating.

V. SELECTED REFERENCES
1. M. Pettenkofer, Ueber ein einfacher verfahren, die dicke einer verzinkung
auf eisen zu schutzen. Munchen, Natl. Tech. Comm. Abh. I, p. 159, 1857.
2. W. A. Patrick and W. H. Walker, Method for testing galvanized iron to re-
place the Preece Test. Ind. Eng. Chem. vol. 3, p. 239, 1911.
3. W. M. Peirce, The Preece Test for zinc-coated materials. Iron Age, vol. 114,
p. 199, 1924.
4. Report of Committee A-5 on Corrosion of Iron and Steel. Proceedings,
Amer. Soc. Testing Matls., vol. 9, p. 100, 1911, and vol. 17, p. 144, 1917.
5. H. Bablik, The Preece Test. Metal Industry (London), vol. 28, p. 369, 1926.
6. Appendix to Standard Methods of determining weight of coating on zinc-
coated articles: Preece or copper sulphate dip test. Amer. Soc. Testing
Matls. Book of A.S.T.M. Standards, part I, p. 323, 1933.
7. S. W. Young and A. E. Stearn, The basic copper sulphates. J. Amer. Chem.
Soc, vol. 38, p. 1947, 1916.
8. H. H. Walkup and E. C. Groesbeck, Some factors affecting the Preece Test
for zinc coatings. Proc. Amer. Soc. Testing Matls., vol. 32, part II, p.
453, 1932.
9. British Standard Specification for the testing of the zinc coating on galvanized
wires, Bull. no. 433, British Standards Institutions, London, 1932.
10. K. H. Kreuchen, Electrolytische untersuchungen an Zinkeinkristallen. Z.
Physik. Chem., vol. 155, Abh. A,, p. 161, 1931.
11. United States Navy Department Specification (22-W-10a), dated June' 1,
1932.
Washington, June 8, 1933.